Assessment of spirometric test quality is important in the interpretation of an individual patient’s study, in determining the overall quality of tests being performed by technologists, and when using the data in research. The 1994 ATS criteria for spirometry require three acceptable trials, two of which are reproducible. Computerized testing equipment often provide a five or six digit error code which summarizes whether these criteria have been met. The system in use at our institution provides an error code that identifies whether the reproducibility criteria have been met, yet reports acceptability only for the designated single best trial. The purpose of this study is to determine how frequently an error code from the best trial that is both acceptable and reproducible accurately indicates that all ATS criteria are met when all trial data are manually reviewed.
All spirometric tests performed on black, white or Hispanic patients in 2003 and 2004 at Beth Israel Medical Center were reviewed. Tests were performed in a SensorMedics Vmax 6200 Autobox (SenorMedics-Viasys, Yorba Linda, CA). The individual trials were manually reviewed for all spirometric studies that had an error code suggesting that ATS criteria for acceptability and reproducibility were met.
There were 2416 tests which had an error code indicating that the best trial was acceptable and that reproducibility criteria were met. Among these 2416 tests, 1926 (79.7%) actually had three or more acceptable trials, 380 (16.0%) had two acceptable trials and 110 (4.6%) had only one acceptable trial.
A final report error code which summarizes whether ATS criteria for reproducibility and acceptability were met can be misleading, unless it also reports how many individual trials meet ATS criteria for acceptability.
Graphical and numerical data, including error codes, should be reviewed from all spirometric trials to best assess test quality prior to the use of the best trial data for interpretation or research.
Mayuko Fukunaga, None.